The formula, available as an online calculator, estimates the child's obesity risk based on its birth weight, the body mass index of the parents, the number of people in the household, the mother's professional status and whether she smoked during pregnancy.

Researchers who published a study of the test in the journal PLOS ONE say they hope it will be used to identify babies at high risk and help families take steps to stop them putting on too much weight before it's too late.

Childhood obesity is a leading cause of early type 2 diabetes, as well as various types of cardiovascular disease, and is becoming increasingly common in developed countries.

According to data from the American Heart Association, almost 18 percent of boys and almost 16 percent of girls aged between two and 19 in the United States are obese. In England, 17 percent of boys and 15 percent of girls aged between two and 15 are obese, according to National Health Service (NHS) data.

"Once a young child becomes obese, it's difficult for them to lose weight - so prevention is the best strategy and it has to begin as early as possible," said Philippe Froguel of Imperial College London, who led the study.

"Unfortunately, public prevention campaigns have been rather ineffective at preventing obesity in school-age children. Teaching parents about the dangers of over-feeding and bad nutritional habits at a young age would be much more effective."

Froguel's team developed the formula using data from a study set up in 1986 following 4,000 children born in Finland.

They were initially investigating whether obesity risk could be assessed using genetic profiles, but the test they developed based on common genetic variations failed to make accurate predictions. Instead, they found that non-genetic information available at the time of birth was enough to predict which children would become obese.

The formula proved accurate not just in the Finnish children they studied, but also in further tests using data from studies in Italy and the United States.

"This test takes very little time, it doesn't require any lab tests and it doesn't cost anything," Froguel said. The calculator is available at:

here

Although the team's work using common genetic variants did not prove to be helpful for predicting childhood obesity, they said about one in 10 cases of obesity are caused by rare mutations that seriously affect appetite regulation.

The researchers said tests for these types of mutations could become available to doctors in the next few years as the cost of DNA sequencing technology falls.

(Reporting by Kate Kelland, editing by Paul Casciato)

• A Permanent Economic Emergency